Stephen -
It's a feat analysis and this is really part of a larger issue. We have 3 groups (schizophrenia, relative, control) and consistently get negative % signal change when looking at the main effects of the condition red/green regardless of the thresholding technique (voxel, cluster, TFCE). Relatives also consistently "outperform" controls.
We want to verify that when we looking at the main effects independently (i.e. just green cue) that we will see only positive activations. We want to make sure there is not a fundamental problem in the data. How can we look only at positive deflections. In an earlier thread I gleaned that if I check the box Z>0 at the group level analysis it will give us only mean activations for the entire group above 0.
Is there a way to verify positive deflections? I currently take the group masked just described and pull the pe from the subject level.
-Edward
|